Intermediate Macroeconomics Lecture 1 - Introduction & Measurement Zs´ ofia L. B´ ar´ any Sciences Po 2014 January
Intermediate MacroeconomicsLecture 1 - Introduction & Measurement
Zsofia L. Barany
Sciences Po
2014 January
About the course I.
I 2-hour lecture every week, Mondays from 17:00-19:00I 3 big topics covered:
1. economic growth2. develop micro-founded framework of a macroeconomic model3. economic fluctuations
I office hour by appointment, in office E.404send me an email if you want to come
I email: [email protected]
I lecture notes will become available weekly on my website
About the course II.
I you will have a problem set related to each lecture
I you will also get the solutions to the problem sets
I you have a tutor: Sylverie Herbert, questions related toproblem sets go to her
I in case many of you have questions related to the same topic,we will organize actual tutorials
What is expected from you?
I attend all the lectures, this is compulsory
I hand in the assigned problem setsin total 50 % of the final grade
I solve all the problem sets at home, this will help you in theexam
I read the additional readings, this will help you develop yourintuition
I take the final exam (after all lectures are finished)
Let’s start!
Overview
I economics: social-economic phenomena are the result ofindividual decision making, and their interactions
I macro vs micro: the differences are in the questions studied,not the methodologyuntil the 1970s the methodology was different as well
I macro: studies the economy as a whole, aims to explainempirically observable phenomenafor example: long run growth, business cycles
I the specific, the individual is so diverse, that even whenidentifying the factswe try to capture the general, common patterns
I to explain we also use simplifications, which capture thegeneral patterns, we create logical constructions, models
Model
I a simplified image of a real phenonemon - captures theessential features of the world needed to analyze a particular(macroeconomic) problem
I a good model can be used for experiments, i.e. to answer thefollowing type of questions: ”what will happen if x changes?”
I model takes the exogenous variables and explains theendogenous variables
I exogenous: something that is given from outside the model
I endogenous: something that is determined within the model
Basic structure
decision makers consumers, firms, government
objectives utility maximization, profit maximization, ?
constraints budget constraint, production technologies, govbudget constraint
I we look at the effect of individual decision making on thegeneral performance of the economy, on aggregate variables
I we define these aggregate variables, analyze their relation toeach other, their evolution over time⇒ identify patterns
I we try to explain why these patterns arise
Output - growth and business cycles
I Gross Domestic Product (GDP): the value of goods andservices produced within a country’s borders over a particularperiod of time
I the time series of the GDP can be separated into trend andbusiness cycle
I trend - in the long run economies tend to grow, long runincrease in the real GDP
I business cycle - the actual real GDP fluctuates around thelong run trendall economies are subject to booms and recessions that occurat irregular time intervals around the trend
US GDP - Raw data (per capita, 2005 dollars)
Clearly an upward trend, but what about the growth rate?
A useful transformation
I The natural logarithm: ln ytI The growth rate of output is gt+1 = yt+1−yt
yt
I We can therefore write yt+1 = (1 + gt+1)yt , and thus
ln yt+1 − ln yt = ln (1 + gt+1)yt − ln yt
= ln(1 + gt+1) + ln yt − ln yt =
= ln(1 + gt+1) ≈ gt+1
I The change in ln(yt) gives therefore the growth rate of ytI The slope of the log-transformed time series, ln yt+1 − ln yt , is
the growth rate of the original series, gt+1
US GDP - Natural Logarithm
Another transformation: Filtering
I A filter (e.g. Hodrick-Prescott filter) decomposes a time seriesinto a business cycle component and a trend
I The trend is economic growth
I The deviation from the trend is the business cycle
US GDP in logs and its trend
Business Cycles - The deviation of US GDP from its trend
Note: These figures are constructed using U.S. data but similarpatterns exist in other OECD countries.
Popular topics in macro
I what causes economic growth? what’s the source ofdifferences over time and across countries?
I can the government influence growth by policy?
I what causes the fluctuations around the trend growth? arethere similarities in the cycles? or are they all different?
I can the government reduce these fluctuations? if yes, shouldthe government intervene?
I unemployment
I credit markets and the financial crisis
I size of the government
I monetary policy, inflation and interest rate
I . . .
Useful Data Sources
I www.ons.gov.uk
I www.ecb.int
I www.oecd.org
I www.worldbank.org
I www.imf.org
I research.stlouisfed.org/fred
Measurement
Topics
I Measuring GDP
I Prices indices – comparing real GDP across time and space
Measurement of GDP
Three ways of measuring the GDP
1. the product approach (value-added approach)
2. the expenditure approach
3. the income approach
Example: Island economy
On this island there is
I a coconut producer
I a restaurant
I consumers
I a government
Island economy: flows
Coconut producer
total revenue $ 20 millionwages $ 5 millioninterest on loan $ 0.5 milliontaxes $ 1.5 million
Restaurant
total revenue $ 30 millionwages $ 4 millioncost of coconuts $ 12 milliontaxes $ 3 million
Consumers
wage income $ 14.5 millioninterest income $ 0.5 milliontaxes $ 1 millionprofits $ 24 million
Government
tax revenue $ 5.5 millionwages $ 5.5 million
1. Product approach
GDP is the sum of value-added across all productive units in theeconomy
value added of coconut producer $ 20 millionvalue added of restaurant $ 18 million (=30-12)value added of government $ 5.5 million
GDP $ 43.5 million
note: the government spends all income on defense expenditure,creating a useful public good by this⇒Value added of government=the cost of inputs to production
2. Expenditure approach
GDP is total spending on all final goods: Y = C + I + G + NX
consumption $ 38 million (=30(rest)+8(coco))investment $ 0 milliongovernment expenditure $ 5.5 millionnet exports $ 0 million
GDP $ 43.5 million
3. Income approach
GDP is the sum of all income received by economic agentscontributing to production: compensation of employees,proprietors’ income, rental income, corporate profits, net interest,indirect business taxes, and depreciation
wage income $ 14.5 millionafter-tax profits $ 24 millioninterest income $ 0.5 milliontaxes (indirect) $ 4.5 million
GDP $ 43.5 million
back to the real world . . .
Composition of U.S. GDP in 2011Components of GDP $Billions % of GDPGDP 15094.0 100
Consumption 10726.0 71.1Durables 1162.9 7.7Nondurables 2483.7 16.4Services 7079.4 46.9
Investment 1916.2 12.7Fixed investment 1870.0 12.4
Nonresidential 1532.5 10.2Residential 337.5 2.2
Inventory investment 46.3 0.3Net Exports -578.7 -3.8
Exports 2085.5 13.8Imports 2664.2 17.7
Government Expenditures 3030.6 20.1Federal defence 824.9 5.5Federal non defence 407.9 2.7State and local gov. 1797.7 11.9
Composition of EU 15 GDP in 2005
Component of GDP $Billions (PPP) % of GDPGDP 11608.6 100
Consumption 6767.4 58.6Investment 2349.9 20.2
Fixed investment 2334.2 20.1Inventory investment 10.8 0.1
Net Exports 65.4 0.6Exports 4186.3 36.1Imports 4120.9 35.5
Government Expenditures 2421.6 20.9
Problems with measuring the GDP
I constructing the GDP in practice is very difficultI there is a trade-off between accuracy and promptnessI based on several data-sources: trend-based data, monthly
data, initial monthly and quarterly data, revised monthly andquarterly data
I at most at quarterly frequencyI several revisions
I GDP is intended as a measure of aggregate output, butI the government sector is priced at input cost (as the output is
not sold in the market)I illegal activities: monetary / non-monetary (i.e. barter)I legal activities: monetary not reported for tax evasion /
avoidance purposes / non-monetary (i.e. barter, DIY,neighbour help)
I Enste and Schneider (2000 JEL) estimated (based onelectricity and currency demand data) that the average size ofthe shadow economy during 1990-93 varies from 12% of GDPfor OECD countries to 39% for developing countries
Interpretation of GDP
I GDP per capita is often used to measure aggregate economicwelfare, but as a welfare measure it leaves out manyimportant factors:
I health, education, environment,. . .I equity: it does not capture the distribution across individuals→ Gini coefficient; percentage of population below the povertyline
I nonmarket activities (but home production time is about thesame as market production time in the Time Use Survey)
I Other attempts to measure welfareI OECD Better Life IndexI UK National Well-being Index (perhaps boosted by Richard
Layard: Happiness - Lessons from a New Science)
GDP and standard of living
Country 2005 GDP per cap life expectancy literacy(2000 int $) (years) (%)
USA 37267 77 97France 30386 80 99UK 29570 79 99Japan 27817 82 99Germany 26210 79 99Hungary 15913 73 99Russia 9648 65 99Mexico 9564 74 92Brazil 7475 69 86China 6012 71 86India 3072 62 60Bangladesh 1827 63 43Nigeria 1003 45 68
Source: nationmaster.com
Comparing GDP
I How to compare GDP across time and across countries?I Differentiating nominal and real GDPI Price indices
Comparing GDP over time
I Some of the change in nominal GDP over time is the result ofchanges in average prices over time
I To measure the real GDP change we need to compare thechange in the volume of goods produced
I Computations involve prices, since GDP comprises manydifferent goods are traded which can only be aggregated invalue terms (we can’t add up apples and oranges, but the $amount spent on them)
I We need a measure of the average prices in the economyI to construct real GDPI to determine how much of an increase in GDP is nominal and
how much is real.
Example: Apples and oranges
apples oranges
quantity year 1 50 100price year 1 $1 $0.8quantity year 2 80 120price year 2 $1.25 $1.6
nominal GDP year 1:
GDP1 = Pa1 · Qa
1 + Po1 · Qo
1 = $1 · 50 + $0.8 · 100 = $130
nominal GDP year 2:
GDP2 = Pa2 · Qa
2 + Po2 · Qo
2 = $1.25 · 80 + $1.6 · 120 = $292
⇒ the nominal GDP increases from $130 to $292. How much isdue to inflation and how much is due to real quantity increases?
Price Indices
I TerminologyI The price level is the average price of all goods and services in
the economyI A price index is a weighted average of the prices of a set of
goods, sometimes called a basket of goodsI Inflation is the change in the price level between two periods
I In practice, changes in the price level, i.e. inflation, areapproximated by the change in a price index
I We can use price indices to transform nominal GDP data intoreal GDP data
I Note: There are many different price indices → differentinflation measures → different real GDP changes
I Two common methodsI Implicit GDP price deflatorI Consumer price index (CPI)
Computation of Real GDPTo compute a real GDP: Fix the prices of goods, and let the basketof goods change
Given a base year, calculate ‘real GDP’ using the base year prices
base yr real GDP year 1 real GDP year 2 realGDPyr1realGDPyr2
year 1 $1 · 50 + $0.8 · 100 80 · $1 + 120 · $0.8 g1 =176130
= $130(=nom. GDP yr 1) = $176 ≈ 1.354%
year 2 $1.25 · 50 + $1.6 · 100 $1.25 · 80 + $1.6 · 120 g2 = 292222.5
=$222.5 = $292(=nom. GDP yr 2) ≈ 1.312%
The real GDP change differs between the two methods! This isdue to the change in the relative price of apples to oranges
Commonly used method: chain-weighting
gc =√
g1 · g2 =√
1.354 · 1.312 = 1.333
if we choose year 1 to be the base year, then year 2 GDP withchain-weighting is: $130 · 1.333 = $173.29
Computation of Real GDPTo compute a real GDP: Fix the prices of goods, and let the basketof goods change
Given a base year, calculate ‘real GDP’ using the base year prices
base yr real GDP year 1 real GDP year 2 realGDPyr1realGDPyr2
year 1 $1 · 50 + $0.8 · 100 80 · $1 + 120 · $0.8 g1 =176130
= $130(=nom. GDP yr 1) = $176 ≈ 1.354%
year 2 $1.25 · 50 + $1.6 · 100 $1.25 · 80 + $1.6 · 120 g2 = 292222.5
=$222.5 = $292(=nom. GDP yr 2) ≈ 1.312%
The real GDP change differs between the two methods! This isdue to the change in the relative price of apples to oranges
Commonly used method: chain-weighting
gc =√
g1 · g2 =√
1.354 · 1.312 = 1.333
if we choose year 1 to be the base year, then year 2 GDP withchain-weighting is: $130 · 1.333 = $173.29
Computation of Real GDPTo compute a real GDP: Fix the prices of goods, and let the basketof goods change
Given a base year, calculate ‘real GDP’ using the base year prices
base yr real GDP year 1 real GDP year 2 realGDPyr1realGDPyr2
year 1 $1 · 50 + $0.8 · 100 80 · $1 + 120 · $0.8 g1 =176130
= $130(=nom. GDP yr 1) = $176 ≈ 1.354%
year 2 $1.25 · 50 + $1.6 · 100 $1.25 · 80 + $1.6 · 120 g2 = 292222.5
=$222.5 = $292(=nom. GDP yr 2) ≈ 1.312%
The real GDP change differs between the two methods! This isdue to the change in the relative price of apples to oranges
Commonly used method: chain-weighting
gc =√
g1 · g2 =√
1.354 · 1.312 = 1.333
if we choose year 1 to be the base year, then year 2 GDP withchain-weighting is: $130 · 1.333 = $173.29
Computation of Real GDPTo compute a real GDP: Fix the prices of goods, and let the basketof goods change
Given a base year, calculate ‘real GDP’ using the base year prices
base yr real GDP year 1 real GDP year 2 realGDPyr1realGDPyr2
year 1 $1 · 50 + $0.8 · 100 80 · $1 + 120 · $0.8 g1 =176130
= $130(=nom. GDP yr 1) = $176 ≈ 1.354%
year 2 $1.25 · 50 + $1.6 · 100 $1.25 · 80 + $1.6 · 120 g2 = 292222.5
=$222.5 = $292(=nom. GDP yr 2) ≈ 1.312%
The real GDP change differs between the two methods! This isdue to the change in the relative price of apples to oranges
Commonly used method: chain-weighting
gc =√
g1 · g2 =√
1.354 · 1.312 = 1.333
if we choose year 1 to be the base year, then year 2 GDP withchain-weighting is: $130 · 1.333 = $173.29
Implicit GDP price deflator
The implicit GDP price deflator is the ratio of nominal to realGDP.There are as many implicit GDP price deflators as there aremethods for computing real GDP.
base year yr 1 yr 2 % change
year 1 1 292/176=1.659 65.9year 2 130/222.5=0.584 1 71.2chain weighted 1 292/173.29=1.685 68.5
The computed rate of inflation (the change in average prices) isdifferent depending on what measure of real GDP we use
Consumer Price Index (CPI)
I Approach the question from another angle
I Fix a basket of goods, and calculate by how much the averagecost of this basket changes over time
I This procedure is also sensitive to which basket we choose
I CPI is the ratio of the cost of the basket (of base yearquantities) in the current year to the cost in base year
I CPI includes only goods and services that are purchased byconsumers.
In our example:
base year CPI in year 1 CPI in year 2 % price change
year 1 1 50·$1.25+100·$1.650·$1+100·$0.8 71.2
year 2 80·$1+120·$0.880·$1.25+120·$1.6 1 65.9
Difference between GDP Deflator and CPI
in our example the inflation rates using the CPI and the implicitGDP price deflator are similar, BUT in practice they can bedifferent
I Conceptual differencesI the GDP deflator is a measure of the prices of all components
of GDP (recall C + I + G + NX ), while the CPI is a measureof only goods bought by the consumers
I the GDP deflator includes only domestic goods, and nothingimported; the CPI includes anything bought by the consumersincluding foreign goods
I These differences might be importantI many wage contracts are tied to the inflation rateI monetary policy aims to stabilise inflation
→ but which one?I the GDP price deflator might give a better measure of inflationI but we might care more about consumers
CPI vs GDP price deflator inflation
Problems in measuring the price level
1. CPI assumes that consumers’ buying habits do not respond toprice changes
example: apples became relatively more expensive, consumersbought relatively fewer apples
⇒ overestimates inflation
2. quality changes
3. new goods
Comparing GDP across countries
I nominal GDP expressed in different currencies
I can compare nominal GDP by using the exchange rate, e, buthow to compare real GDP?
I suppose P is the price level of goods and services in Franceand P∗ is the price level of goods and services in China
I suppose that one Chinese Yuan is worth e euros → e · P∗ isthe price level of Chinese goods and services in euros
I the law of one price states that the price of an identicalgood should be same across countries, i.e. if Pbm is the priceof a BigMac in France, and P∗
bm is the price of a BigMac inChina, then Pbm = e · P∗
bm should hold
I But in practice: the prices of identical goods sold in differentcountries vary significantly, even after we express prices inunits of the same currency
I right question: how much money would be needed topurchase the same goods and services in two countries?
I use these amounts of money to calculate and implicitexchange rate
I → purchasing power parity (PPP) technique, which gives realGDP based on PPP (in international dollars)
I if P = e · P∗, then the purchasing power parity theory wouldhold: the exchange rate (e) would be exactly equivalent tothe implicit exchange rate given by PPP
I Penn effect: goods and services (especially services, which arenon-tradable) tend to be more expensive in richer countries:P > e · P∗
I ⇒ real GDP comparisons across countries by just expressingall prices in the same currencies exaggerate the differences inincome between rich and poor countries
Problems in calculating PPP
1. hard to find a comparable basket of goods
2. prices across countries do not differ in a uniform way: theprice of food may differ less, than the price of housing, orentertainment...
3. the typical basket of goods consumed is also different acrosscountries
4. adjustment for the quality of goods might be necessary as well
5. ... plus all the problems that any price index has
PPP comparisons are still very useful:Penn World Tables make international real GDP comparisonsbased on PPP, see http://pwt.econ.upenn.edu/